Risk assessment study of hydrogen energy storage system based on KPCA-TSO-LSSVM

• Published in: International journal of hydrogen energy 2024-08, Vol.79, p.931-942
• Main Authors: Liu, Jicheng, Song, Yanan, Yu, Xue
• Format: Article
• Language: English
• Subjects: Hydrogen energy storage system; Kernel principal component analysis; Risk assessment; Support vector machine; Tuna swarm optimization algorithm
• ISSN: 0360-3199
• DOI: 10.1016/j.ijhydene.2024.07.070

In order to improve the accuracy and efficiency of hydrogen energy storage system (HESS) risk assessment, the study proposes a risk portfolio assessment model based on kernel principal component analysis-tuna swarm optimization-least squares support vector machine (KPCA-TSO-LSSVM) algorithm. Firstly, the original data were downscaled using KPCA to extract principal components with at least 98% information content. These principal components use the extracted principal components as inputs to the model. Secondly, TSO is used to implement the optimization of parameter settings for LSSVM. Thirdly, the applicability of the proposed KPCA-TSO-LSSVM in HESS risk assessment is verified by case analysis. Finally, the superiority of the model proposed is verified by comparing TSO, whale optimization algorithm (WOA) and particle swarm optimization (PSO). The results show that KPCA-TSO-LSSVM performs optimally in HESS risk assessment. The classification time of the test sample is shorter at 0.0416 s. The accuracy is higher at 97%. Therefore, the proposed model can effectively identify HESS risks, reduce the operational risks of HESS, and improve the stability and reliability of the system. •Kernel principal component analysis extracts principal components of 98% or more.•Tuna swarm algorithm optimises parameters of least squares support vector machine.•Case study validates the applicability of KPCA-TSO-LSSVM in risk assessment.•Multi-method comparisons confirm that KPCA-TSO-LSSVM is the best with 97% accuracy.